impact of fluoroquinolone treatment on delay of tuberculosis diagnosis a systematic review and meta analysis

Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 Contents lists available at ScienceDirect Journal of Clinical Tuberculosis and Other Mycobacterial Diseases journal homepage: www.elsevier.com/locate/jctube Impact of fluoroquinolone treatment on delay of tuberculosis diagnosis: A systematic review and meta-analysis Catherine A Hogan a, Lekha Puri a, Genevieve Gore b, Madhukar Pai a,c,∗ a McGill International Tuberculosis Centre & Department of Epidemiology, Biostatistics and Occupational Health, McGill University 1020 Pine Ave West, Montreal, QC H3A 1A2, Canada b Schulich Library of Science and Engineering, McGill University 809 Sherbrooke St West, Montreal, QC H3A 0C1, Canada c Manipal McGill Center for Infectious Diseases, Manipal University, Manipal, India a r t i c l e i n f o Article history: Received November 2016 Revised December 2016 Accepted December 2016 Keywords: Tuberculosis Fluoroquinolones Diagnostic delay a b s t r a c t Background: Fluoroquinolones are among the most commonly used antibiotics for the treatment of respiratory infections Because fluoroquinolones show bactericidal activity against Mycobacterium tuberculosis, there is concern that their use can delay the diagnosis of tuberculosis We conducted a systematic review and meta-analysis to assess whether empiric treatment with fluoroquinolones delays the diagnosis and treatment of tuberculosis in patients with respiratory tract infections Objectives: The primary objective was to assess the delay in days in the diagnosis and treatment of tuberculosis, among patients who received quinolones, compared to those who received non-fluoroquinolone antibiotics Methods: We included studies of adult patients treated with fluoroquinolones prior to a confirmed diagnosis of tuberculosis We performed a literature search of databases (including PubMed, Embase and Cochrane Library) with no language restrictions We calculated an unweighted mean of estimate of difference in delay across all studies For the studies for which the estimate was available as a mean with standard deviation, a weighted average using a random effects meta-analysis model was estimated Results: A total of 3983 citations were identified from the literature search; of these, 17 articles were selected for full-text review A total of 10 studies were retained for the synthesis These included retrospective cohort studies and case-control studies Only one of these studies was from a high TB burden country, South Africa The most commonly used fluoroquinolones were levofloxacin, gemifloxacin and moxifloxacin The unweighted average of difference in delay between the fluoroquinolone group and non-fluoroquinolone group was 12.9 days (95% CI 6.1–19.7) When these differences were pooled using a random effects model, the weighted estimate was 10.9 days (95% CI 4.2–17.6) When stratified by acid-fast smear status, the delay was consistently greater in the smear-negative group Conclusion: Although results are variable, the use of fluoroquinolones in patients with respiratory infections seems to delay the diagnosis of TB by nearly two weeks Consistent with the International Standards for TB Care, their use should be avoided when tuberculosis is suspected © 2016 The Authors Published by Elsevier Ltd This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Introduction Fluoroquinolones are among the most widely used antibiotics [1] and they are commonly used for the treatment of respiratory tract infections Because fluoroquinolones show bactericidal activity against Mycobacterium tuberculosis complex, there is concern ∗ Corresponding author at Canada Research Chair in Epidemiology & Global Health, McGill University, Dept of Epidemiology & Biostatistics, 1020 Pine Ave West, Montreal, QC H3A 1A2, Canada E-mail address: madhukar.pai@mcgill.ca (M Pai) that their use can delay the diagnosis of tuberculosis The literature on this topic has yielded variable results regarding this association The International Standards for TB Care recommend against the use of fluoroquinolones in adults with suspected tuberculosis [2]; however, these recommendations may not be followed consistently, as shown in recent simulated patient studies from India [3,4] Moreover, both the IDSA/ATS as well as the European guidelines for the treatment of community-acquired pneumonia include respiratory fluoroquinolones (moxifloxacin, gemifloxacin or levofloxacin) as a first-line treatment option for patient with co-morbidities [5, 6] A previous meta-analysis on the topic was published in 2011 [7]; http://dx.doi.org/10.1016/j.jctube.2016.12.001 2405-5794/© 2016 The Authors Published by Elsevier Ltd This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) 2 C.A Hogan et al / Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 Table Study quality assessment of the ten studies included in the systematic review Main author, publication year (reference) Definition of delay∗ Patient sampling Consecutive or random (low risk of bias) Not consecutive or random (high risk of bias) Not described (no description) Clearly stated delay definition (low risk of bias) Rush 2016 [11] All cases included Clearly stated Wang 2015 [12] All cases included Clearly stated Kim 2013 [13] Wang 2011 [14] Not described All cases included Jeon 2011 [15] Not described Definition of confirmation of TB diagnosis Stated but delay definition unclear (high risk of bias) No definition (no description) Clearly stated method of Dx of TB (low risk of bias) No definition (no description) Dealing with confounding Potential confounders identified and adjusted for (low risk of bias) Culture or PCR Confounders identified but not adjusted for Confounders identified but not adjusted for Confounders identified but not adjusted for Confounders identified but not adjusted for Confounders identified but not adjusted for 50–60% by culture or NAAT; rest by ICD code Clearly stated Culture Clearly stated Culture Clearly stated Culture Potential confounders not identified, or identified but not adjusted for (high risk of bias) Wang 2006 [16] All cases included Clearly stated Culture Sierros 2006 [17] All cases included; some cases later excluded re: availability of medical records or results Clearly stated Culture Clearly stated Culture Confounders identified but not adjusted for Clearly stated Culture Confounders identified but not adjusted for Clearly stated Culture Golub 2005 [18] All cases eligible but sampling strategy not described Not described Yoon 2005 [19] Dooley 2002 [20] All cases included Yes, adjusted for smear status Yes, adjusted for smear status Yes, adjusted for smear status ∗ The definition of delay varied between the studies: time of sputum collection to initiation of anti-TB medications: studies; presentation to initiation of anti-TB medications: studies; time of initiation of antibiotics to initiation of anti-TB medications: studies; time of sputum collection to culture growth: study since then, new studies have been published, warranting an updated systematic review Methods We conducted a systematic review and meta-analysis to assess whether empiric treatment with fluoroquinolones delays the diagnosis and treatment of tuberculosis in patients with lower respiratory tract infections The primary objective was to assess the delay in days in the diagnosis and treatment of tuberculosis, among patients who received fluoroquinolones, compared to those who were treated with non-fluoroquinolone antibiotics We aimed to limit selection bias by restricting the comparison to patients who had received nonfluoroquinolone antibiotics only instead of patients who had not received any antibiotics This was done to ensure comparability of both groups as patients may be subject to differential prescribing patterns and medical investigation procedures depending on their pre-test probability of tuberculo-sis as determined by their treating physician We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [8] for the planning and execution of this study With the assistance of a medical librarian (GG), we performed a literature search of seven databases (PubMed, Embase, CINAHL, Cochrane Library, Web of Science, Biosis and Global health) with no language restrictions through to Table Baseline characteristics of the ten studies included in the systematic review Main author, publication year (reference) Setting, country Study design Fluoroquinolone used (number of patients) Antibiotics used in the non-fluoroquinolone group Duration of fluoroquinolone exposure (mean (SD) or median (IQR) in days) Time from presentation to initiation of anti-TB treatment (mean (SD) or median (IQR) in days) All patients Difference between FQ and non-FQ groups (days) AFB smear positive 14 (IQR 3–33) (IQR 1–5) 63.0 (47.1) 12.0 N/A N/A 13.6 N/A N/A 28.0 N/A N/A N/A N/A N/A N/A 15 (IQR 8–175) (IQR 0–105) 32 (no IQR) 42 (IQR 6–233) 27 (IQR 0–198) 46 (no IQR) 18 (no IQR) N/A 34.5 (no IQR) N/A Rush 2016 [11] Inpatients, Canada Retrospective cohort Moxi (9) N/A N/A FQ Wang 2015 [12] Inpatients and outpatients, Taiwan Retrospective cohort Levo (908), moxi (716), cipro (402), oflox (25) Penicillins, cephalosporins, carbapenems, macrolides ≥7 Non-FQ FQ 33 2051 Kim 2013 [13] Inpatients, South Korea Case-control Moxi (10), levo (4), cipro (2) Cephalosporins, macrolides, betalactam/beta-lactamase inhibitor 14 (IQR 7–37) Non-FQ FQ 14,632 16 49.4 (47.6) 35 (IQR 14–72) Non-FQ 32 Wang 2011 [14] Inpatients and outpatients, Canada Retrospective cohort Cipro, Levo, Moxi (N/A) Macrolides, penicillins, sulfonamide, cephalosporins N/A FQ 36 (IQR 3–63) 53.2 (53.6) Jeon 2011 [15] Inpatients and outpatients, South Africa Retrospective cohort Cipro, oflox (N/A) N/A ≥5 Non-FQ FQ 258 90 53.4 (46.7) 23.1 (28.9) 3.3 Wang 2006 [16] Inpatients, Taiwan Retrospective cohort Cipro (42), levo (21), moxi (16) N/A Non-FQ FQ 511 79 19.8 (30.1) 45.4 (37.3) 17.7 Non-FQ 218 27.7 (30.5) FQ 20 N/A Non-FQ 48 N/A FQ 45 39.0 (no IQR) Non-FQ 40 38.5 (no IQR) Sierros 2006 [17] Golub 2005 [18] Inpatients and outpatients, USA Inpatients and outpatients, USA Retrospective cohort Case-control Levo (17), cipro (2), gati (1) N/A Beta-lactam/betalactamase inhibitor, macrolides Macrolides, cephalosporins, amoxicillin, TMP-SMX, clindamycin, vancomycin 9.5 (6.0) (no IQR) N/A 14 (smear-positive) 0.5 AFB smear negative C.A Hogan et al / Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 Number of participants (continued on next page) Main author, publication year (reference) Setting, country Study design Fluoroquinolone used (number of patients) Antibiotics used in the non-fluoroquinolone group Duration of fluoroquinolone exposure (mean (SD) or median (IQR) in days) Time from presentation to initiation of anti-TB treatment (mean (SD) or median (IQR) in days) Number of participants All patients Difference between FQ and non-FQ groups (days) AFB smear positive AFB smear negative Yoon 2005 [19] Inpatients, South Korea Case-control Levo (7), cipro (2) Cephalosporins, macrolides, betalactam/beta-lactamase inhibitor, aminoglycosides 14.2 (8.3) FQ 43.1 (40.0) 24.4 N/A N/A Dooley 2002 [20] Inpatients and outpatients, USA Retrospective cohort Levo (11), trova (2), gati (1), cipro (1), cipro then trova (1) Cephalosporins, macrolides, TMP-SMX N/A Non-FQ FQ 19 16 18.7 (16.9) 21 (IQR 5–32) 16 (no IQR) 24 (no IQR) Non-FQ 17 (IQR 1–16) (no IQR) 16 (no IQR) SD: standard deviation; IQR: inter-quartile range; moxi: moxifloxacin; N/A: not available; FQ: fluoroquinolone; levo: levofloxacin; gati: gatifloxacin; trova: trovafloxacin; cipro: ciprofloxacin; gemi: gemifloxacin; oflox: ofloxacin; USA: United States of America; TMP-SMX: trimethoprim-sulfamethoxazole C.A Hogan et al / Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 Table (continued) C.A Hogan et al / Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 Fig PRISMA flow diagram August 8th 2016 We excluded case series of less than 10 patients, as well as articles for which either the full text or information on diagnostic delay was not available The titles and abstracts of identified articles were screened independently by two reviewers (CH and LP) based on a prespecified protocol Publications identified as relevant by at least one reviewer were searched for full-text appraisal Two reviewers completed full-text evaluations, selected articles, and extracted the data independently Study quality assessment was performed by two reviewers using a tool (Table 1) developed at McGill University for TB diagnostic delay studies [9] Disagreement was resolved by consensus Study authors were contacted in case of ambiguity or missing data, and for clarification of study methodology The data were analyzed with the use of Stata version 12.1 (Statacorp, Texas, USA) The main outcome for the analysis was delay in time of presentation to time of start of anti-tuberculosis therapy in the fluoroquinolone group compared to the non-fluoroquinolone group The threshold for statistical significance was set at a pvalue of 0.05 First, an unweighted average of the estimate of de- lay was calculated for all studies Second, for the studies for which the estimate was available as a mean with standard deviation, a weighted average using a random effects meta-analysis model was estimated Results A total of 3983 citations were identified from the literature search; of these, 17 articles were selected for full-text review A total of 10 studies were retained for the synthesis (Fig 1) These included retrospective cohort studies and case-control studies Only one of the 10 studies was from a high TB burden country, South Africa A total of 18,179 patients were included in the 10 studies The proportion of male participants in the studies ranged from 54 to 99%; HIV seroprevalence ranged from to 58% The most commonly used fluoroquinolones were levofloxacin, gemifloxacin, moxifloxacin and ciprofloxacin For most studies, study quality was considered fair for the definition of delay and for the definition of confirmation of diagnosis C.A Hogan et al / Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 (Table 1) There was limited information available regarding patient sampling methods, as well as strategies to deal with potential confounders In addition, most studies did not adjust their estimates of diagnostic delay by potential confounders The mean difference in delay in time of presentation to time of start of anti-tuberculosis therapy in the fluoroquinolone group compared to the non-fluoroquinolone group varied from 0d to 28d across the 10 studies (Table 2) The unweighted average of difference in delay was 12.9 days (95% CI 6.1–19.7) The estimate of delay was available as a mean with standard deviation for five of the ten studies When the differences for these studies were pooled using a random effects model, the weighted estimate was 10.9 days (95% CI 4.2–17.6) When stratified by acid-fast smear status, the delay was consistently greater in the smear-negative group Based on one study, there was suggestion of a proportional increase in the diagnostic delay based on the duration of fluoroquinolone exposure (18.3 days in the one-day exposure group; 23.1 days in the ≥ day group) However, there was overlap in the confidence intervals around these estimates Heterogeneity in the study design and outcomes was present although this was not statistically evaluated tured In addition, data were lacking on variables such as delay in obtaining a diagnostic procedure (bronchoscopy or sputum induction), and on potential confounders (e.g smear status, previous tuberculosis) Sufficient information was not available to stratify the meta-analysis diagnostic delay estimate by HIV status Furthermore, we attempted to limit bias by restricting the comparison to patients who had received antibiotics only Unfortunately, this restriction was not possible for two studies [17,20] where the groups were mixed with those having received non-fluoroquinolone antibiotics, and those not having received antibiotics This information was not available for one study [15] We recognize that physician behavior may play an important role in accelerating or not the investigations for a patient depending on the underlying suspicion of tuberculosis Finally, the study with the largest number of patients did not include culture diagnosis as necessary for inclusion in the study; this may have decreased power to detect a significant difference between the fluoroquinolone and non-fluoroquinolone groups In addition, the approximation that medians and means were similar may not have been valid in the context of non-normally distributed data Discussion A previous systematic review and meta-analysis on this topic was published in 2011 [7], and reported a mean diagnostic delay of 19.0 days (95% CI 10.9–27.2) This systematic review serves as an update with the addition of six new studies Our mean diagnostic delay of 12.9 days (95% CI 6.1–19.7) also supports fluoroquinolones contributing to the diagnostic delay of tuberculosis However, heterogeneity across studies suggests that this pooled estimate must be interpreted with caution The most plausible mechanism to explain this observed delay relates to the bactericidal activity of fluoroquinolones against Mycobacterium tuberculosis [17], thus delaying the diagnosis of tuberculosis Moreover, fluoroquinolones may transiently improve the respiratory symptoms of tuberculosis, also contributing to delay In vitro experiments have demonstrated that the bactericidal power of moxifloxacin and levofloxacin against Mycobacterium tuberculosis far outweighs that of ciprofloxacin [10] The potent bactericidal activity of moxifloxacin in patients with pulmonary tuberculosis has made it a candidate of choice for the new short-course studied regimens, including PAMZ (pretomanid, moxifloxacin and pyrazinamide; currently in phase (3) and BPaMZ (bedaquiline, pretomanid, moxifloxacin and pyrazinamide; currently in advanced phase (2) As such, there is a compelling drive to protect quinolones from possible emerging resistance, an outcome this study did not assess The suggestion of a proportional relationship between the duration of fluoroquinolone exposure and delay in tuberculosis diagnosis is interesting and deserves further attention The only study [15] that reported on the diagnostic delay of tuberculosis as a function of time categorized fluoroquinolone use by day, 2–4 days and ≥ days There was a trend toward longer diagnostic delay in the group treated for ≥ days compared to the two other groups; however, the confidence intervals overlapped around these estimates This suggests that there may be a benefit in stopping the fluoroquinolone at the beginning of treatment if one thinks of tuberculosis as a diagnostic possibility even after the start of empiric antibacterial treatment Our study presents certain limitations In particular, there was important variation in study settings, patient populations and in the definitions of delay used for the studies Generalizability may also be limited given that only one study was from a high TB incidence country Most studies reported delay as defined by the time to initiation of treatment; as such, the extra time between time of diagnosis and time of initiation of treatment is not adequately cap- Conclusion Although the available literature has several limitations and study results are heterogeneous, the evidence suggests that the use of fluoroquinolones in patients with respiratory infections might delay the diagnosis of active pulmonary TB by nearly weeks Consistent with the International Standards for TB Care, their use should be avoided when tuberculosis is suspected Conflicts of interest None of the authors have any financial or industry conflicts MP serves as a consultant to the Bill & Melinda Gates Foundation (BMGF) BMGF had no involvement this manuscript Funding MP is supported by a Canada Research Chair award This agency had no involvement in this manuscript References [1] Center for Disease Dynamics, Economics and Policy State of the world’s antibiotics, 2015 Washington, D.C.: CDDEP; 2015 [2] Tuberculosis Coalition for Technical Assistance International standards for tuberculosis care (ISTC) The Hague: Tuberculosis Coalition for Technical Assistance, 2006 [3] Das J, Kwan A, Daniels B, Satyanarayana S, Subbaraman R, Bergkvist S, et al Use of standardised patients to assess quality of tuberculosis care: a pilot, cross-sectional study Lancet Infect Dis 2015;15(November(11)):1305–13 doi:10.1016/S1473-3099(15)0 077-8 [4] Satyanarayana S, Kwan A, Daniels B, Subbaraman R, McDowell A, Bergkvist S, et al Use of standardised patients to assess antibiotic dispensing for tuberculosis by pharmacies in urban India: a cross-sectional study Lancet Infect Dis 2016 Aug 24 doi:10.1016/S1473-3099(16)30215-8 [5] Mandell LA, Wunderink RG, Anzueto A, Bartlett JG, Campbell GD, Dean NC, et al Infectious Diseases Society of America; American Thoracic Society Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults Clin Infect Dis 2007;44(Suppl 2):S27–72 Mar [6] Woodhead M, Blasi F, Ewig S, Garau J, Huchon G, Ieven M, et al Joint Taskforce of the European respiratory society and European Society for clinical microbiology and infectious diseases.Guidelines for the management of adult lower respiratory tract infections–full version Clin Microbiol Infect 2011;17(November (Suppl 6)):E1–59 [7] Chen TC, Lu PL, Lin CY, Lin WR, Chen YH Fluoroquinolones are associated with delayed treatment and resistance in tuberculosis: a systematic review and meta-analysis Int J Infect Dis 2011;15(March(3)):e211–16 doi:10.1016/j ijid.2010.11.008 C.A Hogan et al / Journal of Clinical Tuberculosis and Other Mycobacterial Diseases (2017) 1–7 [8] Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement PLoS Med 20 09;6(7):e10 0 097 doi:10.1371/journal.pmed.10 0 097 [9] Qin, Z.Z Delays in diagnosis and treatment of pulmonary tuberculosis, and patient care-seeking pathways in China: a systematic review and meta-analysis Master’s thesis 2015 [10] Cremades R, Rodríguez JC, García-Pachón E, Galiana A, Ruiz-García M, López P, et al Comparison of the bactericidal activity of various fluoroquinolones against Mycobacterium tuberculosis in an in vitro experimental model J Antimicrob Chemother 2011;66(October(10)):2281–3 doi:10.1093/jac/dkr281 [11] Rush B, Wormsbecker A, Stenstrom R, Kassen B Moxifloxacin use and its association on the diagnosis of pulmonary tuberculosis in an inner city emergency department J Emerg Med 2016;50(March(3)):371–5 doi:10.1016/j.jemermed [12] Wang JY, Lee CH, Yu MC, Lee MC, Lee LN, Wang JT Fluoroquinolone use delays tuberculosis treatment despite immediate mycobacteriology study Eur Respir J 2015;46(August(2)):567–70 doi:10.1183/09031936.0 019915 [13] Kim SY, Yim JJ, Park JS, Park SS, Heo EY, Lee CH, et al Clinical effects of gemifloxacin on the delay of tuberculosis treatment J Korean Med Sci 2013;28(March(3)):378–82 doi:10.3346/jkms.2013.28.3.378 [14] Wang M, Fitzgerald JM, Richardson K, Marra CA, Cook VJ, Hajek J, et al Is the delay in diagnosis of pulmonary tuberculosis related to exposure to fluoroquinolones or any antibiotic? Int J Tuberc Lung Dis 2011;15(August(8)):1062–8 doi:10.5588/ijtld.10.0734 [15] Jeon CY, Calver AD, Victor TC, Warren RM, Shin SS, Murray MB Use of fluoroquinolone antibiotics leads to tuberculosis treatment delay in a South African gold mining community Int J Tuberc Lung Dis 2011;15(January(1)):77–83 [16] Wang JY, Hsueh PR, Jan IS, Lee LN, Liaw YS, Yang PC, et al Empirical treatment with a fluoroquinolone delays the treatment for tuberculosis and is associated with a poor prognosis in endemic areas Thorax 2006;61(October(10)):903–8 doi:10.1136/thx.2005.056887 [17] Sierros V, Khan R, Joo Lee H, Sabayev V The effect of fluoroquinolones on the acid-fast bacillus smear and culture of patients with pulmonary tuberculosis Clin Pulm Med 2006;13(May(3)):164–8 doi:10.1097/01.cpm.0000218459.07407 1d [18] Golub JE, Bur S, Cronin WA, Gange S, Sterling TR, Oden B, et al Impact of empiric antibiotics and chest radiograph on delays in the diagnosis of tuberculosis Int J Tuberc Lung Dis 2005;9(April(4)):392–7 [19] Yoon YS, Lee HJ, Yoon HI, Yoo CG, Kim YW, Han SK, et al Impact of fluoroquinolones on the diagnosis of pulmonary tuberculosis initially treated as bacterial pneumonia Int J Tuberc Lung Dis 2005;9(November(11)):1215–19 [20] Dooley KE, Golub J, Goes FS, Merz WG, Sterling TR Empiric treatment of community-acquired pneumonia with fluoroquinolones, and delays in the treatment of tuberculosis Clin Infect Dis 2002;34(12):1607–12 Jun 15 doi:10 1086/340618